This project will continue to focus on the identification of environmental risk factors for hepatocellular cancer (HCC) and the development of early biomarkers of disease. In ongoing studies, we demonstrated that, in conjunction with hepatitis B or C virus infection, elevated levels of biomarkers of aflatoxin B1 (AFB1) and polycyclic aromatic hydrocarbon (PAH) exposure and oxidative stress are associated with increased HCC risk in Taiwan. These studies are using cases and controls nested in the Cancer Screening Program (CSP) cohort established by our collaborator Dr. Chien-Jen Chen in Taiwan. A total of almost 24,000 subjects were recruited between 1990-1991 and by the end of the next grant period (July 2014), at least 700 HCC cases will have been identified. We will continue to determine levels of several biomarkers (AFB1- and PAH-albumin, urinary aflatoxin metabolites, 8-oxo-guanine and 15-F2t-isoprostane) in baseline samples from newly identified cases and controls in the CSP cohort. The larger sample size will provide more statistical power to test new hypotheses related to methylation and HCC risk. Repeat blood samples from this cohort have allowed us to demonstrate that gene-specific methylation of tumor DNA circulating in blood can be detected years prior to clinical diagnosis and that the frequency of methylation is significantly higher in those who go on to develop HCC than in those who do not. While these studies analyzed one candidate gene at a time, we now propose to use Illumina arrays interrogating 27,000 CpG sites to analyze liver tissues to more broadly select appropriate genes to improve screening. We will use our sample bank to assay paired tumor/adjacent tissues from Taiwan but also expand these studies to US HCC cases. Genes determined to be differentially methylated will be analyzed in DNA from plasma collected in the CSP cohort from cases prior to diagnosis and in controls to determine the predictive value of the methylation markers. We will use the same approach in US cases to determine whether plasma DNA methylation predicts recurrence in those undergoing surgery for HCC. We will also explore whether WBC DNA gene-specific methylation of these genes in baseline bloods predicts HCC risk within the CSP cohort. In addition to gene-specific methylation, it is now known that global methylation levels are decreased (hypomethylation) in tumors and several studies have now demonstrated decreased WBC DNA methylation in cases compared to controls. Thus, we propose to also investigate global methylation in tissue and WBC DNA, determining their correlation. We will determine whether global methylation is decreased in tumor compared to adjacent tissues and decreased in WBC DNA of cases compared to controls. We will also determine the correlations of methylation to biomarkers of environmental exposures. These studies will help in the identification of high risk individuals and development of screening methods.